(PS6-C61) Identifying Subgroups of Cannabis Users from Expectancies and Use Indices: A Latent Profile Analysis

College students maintain elevated prevalence rates of cannabis use and associated increased risk for negative cannabis outcomes. Extant research has elucidated variables associated with cannabis use and negative cannabis outcomes, such as cannabis expectancies (i.e., expectations about the expected effects of cannabis) and motives (i.e., specific reasons for use). Motivational models of cannabis use posit that expectancies serve to incentivize cannabis use based on perceived instrumental effects of cannabis. Further, cannabis motives are posited as the final common pathway to cannabis use, such that motives mediate relations between distal factors (e.g., expectancies) and use. Thus, understanding relations between cognitive mechanisms underlying use and cannabis outcomes may inform the maintenance of cannabis use, and identification of relevant risk factors. However, extant research in this area has typically employed variable-centered analytic approaches, which may fail to appreciate heterogeneity of cannabis users’ expectancies. To address this limitation, the present study utilized person-centered analysis to identify subgroups of cannabis users based on expectancies and cannabis use behaviors. A multisite sample of college students (N = 1171) endorsing past-month cannabis use completed questionnaires regarding their cannabis expectancies, motives, past-month use, past-month negative consequences, and risk for cannabis use disorder (CUD). Latent profile analysis (LPA) was employed, such that expectancy subscales, cannabis use frequency, and cannabis use quantity served as indicators. Subsequently, identified groups were compared for differences across cannabis use consequences, risk for CUD, and motives using the Bolck-Croon-Hagenaars (BCH) method. LPA identified six unique profiles: (1) “Low use/expectancies” (7.0%), (2) “Low-moderate use/low-moderate expectancies” (17.6%), (3) “High use/moderate expectancies” (22.5%), (4) “Low use/moderate expectancies” (26.4%), (5) “Low-moderate use/moderate expectancies” (19.8%), and (6) “Moderate use/high expectancies” (6.7%). Omnibus tests indicated differences across latent profiles for negative consequences (χ²(5) = 309.45, p < .001), risk for CUD (χ²(5) = 659.10, p < .001), and all cannabis motives (all p < .001). Generally, groups with elevated use and expectancies tended to report higher negative consequences, risk for CUD, and motives. Moreover, profile three reported experiencing fewer negative consequences (χ²(1) = 17.35, p < .001) and lower risk for CUD (χ²(1) = 24.58, p < .001) relative to profile six, despite heavier cannabis use. This supports the utility of considering indices beyond cannabis use (e.g., expectancies) when predicting negative cannabis outcomes. Overall, these findings document heterogenous groups of cannabis users based on expectancies and use patterns, while also supporting previous findings on relations between expectancies, motives, use, and negative outcomes. Future research is needed to replicate the validity of the identified profiles. However, these findings attest to the potential utility of using expectancy profiles as a means to assess relative risk for negative cannabis outcomes.